Part of the problem here is obviously the lack of a small, localizable data set that we can use to consistently reproduce the problem to be able to determine if it is:
* An issue with the feature data that the KingOracle FDO provider or MapGuide is failing to properly process/sanitize. * A threading issue with the KingOracle FDO provider. Try using a tool like gatling (http://gatling.io) and write/record a "user [selects feature -OR- hovers for tooltips] on oracle layer" scenario and then see how easy it is to crash with the same error as you crank up the repetitiveness and concurrent users when replaying that particular scenario. Then if you can get this scenario to consistently crash, capture as much data on the requests that cause the crash (the crashing request will most likely return a http status code of 559, so you could programmatically save the request parameters of any request whose response is 559). Then with this data, you can try to look for patterns in the offending requests (ie. Does it consistently crash on a particular input geographic polygon, or a particular combination of QUERYMAPFEATURES parameter values?) At this point, there should hopefully be enough data points to be able to determine if this is a problem in how the KingOracle FDO provider is reading the feature data or a thread safety issue in the KingOracle FDO provider, and be able to take further action from there. - Jackie -- View this message in context: http://osgeo-org.1560.x6.nabble.com/AIMS-crashing-with-King-Oracle-FDO-layers-and-multiple-users-tp5256689p5256749.html Sent from the MapGuide Users mailing list archive at Nabble.com. _______________________________________________ mapguide-users mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/mapguide-users
